
*********************************************************************
*	On What Moves (Spending) Mood?                                  *
*   The Nature and Origins of Parallel Public Preferences           *
*                                                                   *
* 	Steven M. Van Hauwaert, ESPOL-LAB & Radboud University          *
*	Christopher Wlezien, University of Texas at Austin              *
* 	Ryan E. Carlin, Georgia State University                        *
*********************************************************************

use "/[pathway]/what moves mood non-panelled 20250513.dta"


****************************************************
// FIGURE 1: PUBLIC OPINION MEASURES //
****************************************************


graph twoway (line pref_env_raw year, sort) (line pref_edu_raw year, sort) (line pref_heal_raw year, sort) (line pref_welf_raw year, sort) if year < 2020, scheme(plotplain) legend(position(6)) saving(preferences)


graph twoway (line treal_APP_env year, sort) (line treal_APP_edu year, sort) (line treal_APP_health year, sort) (line treal_APP_welfare year, sort) if year < 2020, scheme(plotplain) legend(position(6)) saving(spending)

graph combine "preferences.gph" "spending.gph", scheme(plotplain)


****************************************************
// TABLE 1: PROPERTIES OF MASS PREFERENCES //
****************************************************


pwcorr pref_env_raw pref_edu_raw pref_heal_raw pref_welf_raw if year < 2020

factor pref_env_raw pref_edu_raw pref_heal_raw pref_welf_raw if year < 2020


****************************************************
// TABLE 6: AGGREGATED PREFERENCES //
****************************************************


* 1) Average social preferences

regress pref_soc_int l.pref_soc_int treal_APP_total gdp_trill president if year < 2020


* 2) Estimate mood from CFA

regress mood_int l.mood_int treal_APP_total gdp_trill president if year < 2020



****************************************************

// SUPPLEMENTARY MATERIALS //

****************************************************


****************************************************
// A. TIME-SERIES DIAGNOSTICS //
****************************************************


// Table A5 //


dfuller gdp_trill if year<2020, regr trend lags(1)
dfuller treal_APP_total if year<2020, regr trend 


// Table A6 //


kpss gdp_trill if year<2020, auto
kpss treal_APP_total if year<2020, auto


// Table A7 //


regr treal_APP_total gdp_trill if year<2020
predict Rtreal, residuals
regr d.Rtreal l.Rtreal if year<2020


****************************************************
// D. ALTERNATIVE MODELS OF NET PREFERENCES BY DOMAIN //
****************************************************


// TABLE D.1

sureg (pref_edu_int l.pref_edu_int treal_APP_total gdp_trill president) (pref_env_int l.pref_env_int treal_APP_total gdp_trill president) (pref_heal_int l.pref_heal_int treal_APP_total gdp_trill president) (pref_welf_int l.pref_welf_int treal_APP_total gdp_trill president) if year < 2020


****************************************************
// F. MODELS EXCL. PARTY OF THE PRESIDENT //
****************************************************


// Table F.3

regress pref_soc_int l.pref_soc_int treal_APP_total gdp_trill if year < 2020

regress mood_int l.mood_int treal_APP_total gdp_trill if year < 2020
